In article <LILJEBER.92Aug17203359@klaava.Helsinki.FI> liljeber@klaava.Helsinki.FI (Mika Pekka Liljeberg) writes:
quoted text
> Would this break demand loading, then? I think not. A text page could still
> be freed and reloaded on demand. Any library references on it would simply
> get relinked, when the path of execution intercepted them.

... unless that address reference happens to cross a page boundary. Then
you have five choices:

   - define this case as illegal. (Not that bad, but probbaly unpopular.)
   - provide a mechanism to fault in both pages at the same time.
     (Might be difficult.)
   - lock both pages in memory. (Very ugly.)
   - make both pages part of the process' data space. (Ugly.)
   - don't patch the call. (Now you're possibly incurring a huge performance
     penalty (at least compared to what you gain by modifying jumps),
     because you're going through the patching code whenever that call is
     performed.)

- Werner
-- 
   _________________________________________________________________________
  / Werner Almesberger, ETH Zuerich, CH      almesber@nessie.cs.id.ethz.ch /
 / IFW A44  Tel. +41 1 254 7213                 almesberger@rzvax.ethz.ch /
/_BITNET:_ALMESBER@CZHETH5A__HEPNET/CHADNET:_[20579::]57414::ALMESBERGER_/